RELATION OF THE PHYSICAL AND CHEMICAL PROPERTIES OF THE SOIL TO THE WILT.

The second line of attack in searching for a practical method of controlling the wilt was to determine whether the soil could be treated by chemical or physical methods which would destroy the organism causing the trouble or weaken its activity. In 1904 a series of field plats on infested soil at Creedmoor, N. C., was given various chemical treatments before being set to tobacco. Heavy applications of powerful disinfectants, including several salts of copper, formalin, corrosive sublimate, permanganate of potash, carbolic acid, iron sulphate, and sulphur, were tried. Strong acids and alkalis and excessive applications of potash, nitrogen, and phosphoric acid also were tested. None of the treatments gave any promise of success in practical tobacco culture. In 1910 another series of field tests with chemicals was made on a diseased field near Creedmoor, including a more extensive study of the effect of acid and alkaline conditions on the disease. The effects on the wilt of soluble forms of calcium, magnesium, silicon, aluminum, iron, and manganese were tested, but without results, these tests being based on the assumption that the wilt is not to be feared in soils containing large amounts of clay. Various fertilizer treatments were tried, only materials tending to produce alkaline conditions being used on some plats, while on others only substances favoring an acid reaction were employed. For example, one plat received per acre 200 pounds of carbonate of potash, 600 pounds of basic slag, 250 pounds of nitrate of soda, 500 pounds of cottonseed meal, and 2,000 pounds of burned lime, while another received equivalent amounts of sulphate of potash, acid phosphate, ammonium sulphate, and acid sodium sulphate. None of these treatments produced any decided effect on the amount of wilt.

In the following year, in order better to control the experimental conditions, a large number of glazed-tile cylinders 2 feet long and 2 feet in diameter were set in the ground so as to leave about 3 inches projecting above the ground level. The cylinders were installed at West Raleigh and at Creedmoor. The pots were filled to a depth of about a foot with bottom earth taken from the holes in which the pots were placed, after which about 8 inches of subsoil and 8 inches of topsoil from a diseased tobacco field at Creedmoor were added. All tests were made in duplicate and each cylinder contained four plants. In continuation of previous work, tests were made with mixtures of varying proportions of sandy and clay soils; with ordinary wilt soil, to which were added separately the chief chemical constituents of clay; with plants grown from seeds without transplant

ing, so as to avoid breaking the roots; with acid reagents and alkaline reagents; and with various chemical disinfectants. Considerable difficulty was experienced in maintaining the infection in the cylinders, because of their becoming excessively dry or from some other unfavorable condition, so the results were rather inconstant and uncertain. The results of tests covering three years, mixtures of sandy and clay soils being used, seemed to indicate that the wilt organism is not very active in clayey soils, but none of the chemical constituents of clays applied singly to sandy soils appeared to have any decided action on the organism. A rather extensive series of tests was made with acids and alkalis. For the former sulphuric acid, acid potassium sulphate, and nitric acid were used, and for the latter quicklime, calcium carbonate, potassium carbonate, sodium silicate, and sodium carbonate were employed. In most cases at least two different rates of application were made. The results of tests covering three years indicate in general that alkalis are favorable to the wilt organism, while acids retard its development, but the results were not very constant and failed to furnish any basis for a practical remedy. Details of the tests are shown in part in Table I. Each cylinder should have contained four plants, but it will be noted that the stand was poor, due largely to the toxic action of the chemicals applied to the soil.

TABLE I.-Tests of tobacco plants grown in cylinders containing soil treated with acids and with alkalis at West Raleigh and Creedmoor, N. C., in 1913 and 1914.

1 Tests located at West Raleigh, except as otherwise stated.

2 Cylinder No. 38 received 4 ounces of nitric acid and cylinder No. 43 received no acid in 1914.

In the cylinder tests with disinfectants, use was made of formalin, potassium permanganate, chlorid of lime, atomic or superfine sulphur, ẞ-naphthol, Bordeaux mixture, and a number of proprietary preparations, mostly coal-tar products, many of them in varying rates of application. No decidedly beneficial results were obtained.

During the four years, 1911 to 1914, inclusive, field plats at Creedmoor, N. C., consisting of single rows, were used in trials with disinfectants. The usual fertilizing and cultural methods were followed. alike on all plats. The results for 1911 and 1912 are shown in part in Table II. The permanganate of potash and chlorid of lime were applied in the same quantities and in double the quantities recommended by Honing in Sumatra, the weaker solutions consisting of 125 grams of the chemical in 25 gallons of water. The solutions were poured into the holes in which the plants were to be set 24 hours. before transplanting.

TABLE II.-Field experiments with tobacco plants, showing the effects of soil disinfectants at Creedmoor, N. C., in 1911 and 1912.

In 1913 and 1914 the experiments were continued along the same lines, with the use of formalin, sulphuric acid, naphthol, carbolic acid, Bordeaux mixture, atomic sulphur, lime sulphur, coal tar, pine tar, and several proprietary disinfectants. In these tests more than a score of germicides and other chemicals were applied in various quantities and in various ways. Since the same plat was used each year, the cumulative effect of the chemicals was detrimental to the growth of the tobacco. Here, as in the cylinders, it was difficult to secure a good stand of plants. In general, there have been no consistent differences in the percentages of wilt in the treated and the untreated rows. The results are confirmatory of the cylinder tests and do not promise success in wilt control through the use of chemicals. In 1915 certain materials supposed to contain radium, including Banque du Radium, carnotite, and a radioactive earth, were tried in several differing quantities at Creedmoor, but without any definite results.

94919°-Bull. 562-17--2

It should be added finally that in view of the theory entertained by some that lack of drainage or aeration in the soil is a cause of wilt, experiments were made on the effects of deep plowing and subsoiling, but the results were entirely negative. Dynamiting the soil according to methods recommended was tried, with a view to destroying any hardpan or impervious layer beneath the topsoil, but no beneficial action could be noted.

CROP ROTATION AS A BASIS FOR THE CONTROL OF THE WILT.

It has been seen that so far as known all species and varieties of. tobacco, both wild and cultivated, are readily destroyed by the wilt, and it is well known that other plants belonging to the same family are seriously attacked. Obviously none of these plants should be grown on tobacco land affected with wilt. The parasite, however, is by no means limited to solanaceous plants, and is now known to attack members of no less than nine distinct families of the higher plants, namely, (1) Solanaceae (including tobacco, Irish potato, pepper, eggplant, jimson weed, etc.), (2) Leguminosa (including the peanut), (3) Balsaminaceæ, (4) Compositæ (including the common ragweed), (5) Euphorbiaceæ, (6) Pedaliacea, (7) Tropaeolacea, (8) Urticaceæ, and (9) Verbenacea. It is probable that further search will show that still other plants are attacked by this organism. Outside of the Solanaceae the peanut and the ordinary ragweed are of special importance in the flue-cured district. In parts of the district peanuts are an important money crop and the ragweed (Ambrosia artemisiafolia) is extremely common on tobacco lands throughout the district. Peanuts must be avoided and the ragweed kept down if rotation of crops is to be effective in controlling the wilt.

CROP-ROTATION EXPERIMENTS AT CREEDMOOR, N. C.

Beginning with 1911, experiments on the effectiveness of crop rotation for the control of tobacco wilt have been conducted on a tobacco field near Creedmoor, N. C., on which tobacco was practically a total failure in 1910. The soil in this field is the Granville coarse sandy loam and is rather uniform throughout except for a gall spot of several square yards on the plat designated "F" in figure 1. The plats slope away gently on either side of the turn row, which extends through the center of the field. The drainage is good on all plats except for small areas on the southern edge of plat 1 and on the northern edge of plat 4, and there is but little chance for surface drainage from one plat to adjoining ones.

In these experiments the rotation and cropping systems included nearly all of the more important crops adapted to the section, namely,

FIG. 1.-Plan of the experiment field at Creedmoor, N. C., showing details of the cropping system on each plat. Plats B, C, D, E, and F were planted to tobacco in 1916.

tobacco, corn, cotton, peanuts, wheat, sweet potatoes, cowpeas, mixed clovers, and grasses. Two different types of cropping were followed. In the first type a single crop or group of crops was grown continuously on the same plat. Thus, corn was grown continuously on a quarter-acre plat for five years, with crimson clover as a winter cover crop. In the same way, peanuts and sweet potatoes were each grown continuously on separate plats for five years, in each case with rye as a winter cover crop. Wheat, followed by cowpeas cut for hay, also was grown continuously on one plat. On one plat a mixture of mammoth or sapling clover, redtop, orchard grass, and tall meadow oat-grass was grown each year, the plat having been reseeded in 1913. A plat cropped continuously to cotton was started two years later and thus far has run only three years. For comparison, tobacco was grown each year on one plat, with rye and crimson clover as a winter crop. In 1916 tobacco was grown on all plats to determine the effects of the other crops on the wilt. There were two objects in the continuous cropping to the same crop plants each year. In the first place it was not certain at the outset which of the crops are immune to the wilt, and, secondly, it was desired to determine by direct test the comparative values of the different crops in reducing the amount of wilt. This plan was fully justified by later development, for it was soon found that peanuts are quite susceptible to the disease, while, on the other hand, the remaining crops proved to be about equally effective in reducing the injury from the wilt, thus indicating that they are immune.

In the second type of cropping, a system of rotation including corn, wheat and cowpeas, red clover and grass, and tobacco was followed on a series of plats. In this series it was arranged that after